Why are balance patches so rare?

[DanteZero.9736]

@starlinvf.1358 said:

@DanteZero.9736 said:I remember an old dev post from early in GW2's release that mentioned Anet was aiming for balance patches that don't cause instant overnight radical shifts in the meta/strategies but rather, slowly tone down imbalanced skills and strategies.

EXAMPLE

A class is dealing say, 45k DPS in PVE and Anet wants to drop it down to 35k, so they release a series of balance patches over the course of a year by making a tweak to each contributing factor in each patch.

The first patch might have two traits nerfed by 5% each, in another patch 3 months later, one or two key skills might have their activation times or cooldown times increased or amount of hits they do reduced by a bit, etc. Repeat again for a third or even a fourth patch.

By the end of the year, the class will have been reduced to 35k DPS in PVE and fans might be furious about it, but they aren't up in arms like they used to be in GW1's balance patches.

Compare that to GW1, where a build would get all their key skills nerfed all at once and cause a lot of angry people to scream about it.

That whole spiel got thrown out the window less then 6 months after that statement, as they made sweeping changes to multiple skills that caused a massive shift that side lined half the meta in one night....... because Rev would love to know what this shaving approach is like.

I'm not trying to call you out or anything malicious, but do you remember a particular case or two? I don't remember that well what those changes are and in what mode you're talking about.

[Bladestrom.6425]

the fact that the balancing patches are so slow suggests that their development pipeline is pretty immature and thats why it takes so long to make changes. For example at the far end of the spectrum there are development companies that have the ability to make hundreds of live deployments a day, the principle is that you have automated testing throughout your full development pipeline, and you make tiny little incremental changes that worst case you could roll back in an hour. So for balancing a game, all they have to do is add a tiny incremental buff to skills at a time and then monitor the metrics. the game meta is not going to collapse for example if a dragons tooth damage went up from say 6000 to 6010. What doesnt work is big bang balance changes that are months appart, thats a pretty out of date development model unless you happen to be designing systems for critical medical apps etc etc

[OriOri.8724]

@vesica tempestas.1563 Continuous integration systems and balance patches are about as far away from each other as you can get while still calling it programming. You don't balance classes by making changes and making sure none of your test cases break. You balance them by toning down what is overpowered and buffing what is underpowered. A continuous integration system has nothing to do with that. Also, in your comment you conveniently ignore all of the other updates that come to the game besides balance patches. All in all pretty much your entire comment is just inaccurate

[zealex.9410]

@Illconceived Was Na.9781 said:

@zealex.9410 said:As for the comments "the community needs to see how stuff works and how balance patches affect the game" they can do so in less than 2 months.

It's a fact that the majority of the community doesn't adapt as fast as Quantify and other top players do. ANet has said it takes time for them to collect the data, identify issues, and then it also takes time to come up with a plan for how to address it, and adjust as needed. None of us can say how long it should take unless we get access to the same data that ANet is using, but we can say, based on ANet's comments that ANet doesn't think it can be done within 2 months (or at least, not for a full balance pass).

During that timewindow of 3 months classes can literally be useless in pvp wvw and pve no moter how much time they need to collect data its clearly ineficient when the same overpowered build remains overpowere for more than 4 balance patches. While other underpowered build remain ded in the water for an equal amount of balance patches.

[zealex.9410]

@DanteZero.9736 said:I remember an old dev post from early in GW2's release that mentioned Anet was aiming for balance patches that don't cause instant overnight radical shifts in the meta/strategies but rather, slowly tone down imbalanced skills and strategies.

EXAMPLEA class is dealing say, 45k DPS in PVE and Anet wants to drop it down to 35k, so they release a series of balance patches over the course of a year by making a tweak to each contributing factor in each patch.

The first patch might have two traits nerfed by 5% each, in another patch 3 months later, one or two key skills might have their activation times or cooldown times increased or amount of hits they do reduced by a bit, etc. Repeat again for a third or even a fourth patch.

By the end of the year, the class will have been reduced to 35k DPS in PVE and fans might be furious about it, but they aren't up in arms like they used to be in GW1's balance patches.

Compare that to GW1, where a build would get all their key skills nerfed all at once and cause a lot of angry people to scream about it.

In gw1 balance happened more often than 1 every 3 months. And no if anet wants to nerf something they will gut it. The year u are talking about is them buffing other aspecs of the class they originally wanted to be played to those same standarts.

[tigirius.9014]

@onevstheworld.2419 said:They do have a balance team.

Balance patches usually happen in the PvP off season to keep it fair. The current season will only finish on the 16th, so don't expect anything until afterwards.

When the game was launched we were told that the damage algorithms for pvp and pve were completely separate so keeping all of the cooldowns and cast times the same wouldn't effect pvp if they changed the damage output of PVE classes. Particularly the nerfed before it was live elite spec Deadeye for example- which works great in PVP but in PVE it's terrible!

[Etterwyn.5263]

@Arzurag.7506 said:

@Oldirtbeard.9834 said:I'm a Power Reaper, getting attention from this balance team seems like a school nurse saving me from a gunshot wound with a band aide with cartoon characters on it while telling me to take a nap on the cot in the back of the Nurse's office.

I know that feeling. Been playing Power-Reaper since I began playing this game 2 years ago and never shifted to conditions.I hope Anet has a few plans for us in their backhand.

10% buff to axe auto. :#

[Llethander.3972]

Outside of the obvious PvP seasons one needs to consider that they need allow time for information able to be gathered after changes are made to begin to average out. If you take early information you risk working based off of incredibly skewed data due to an insufficient sample size of relevant results.

Essentially it's the difference between flipping a coin four times and taking the average numbers to work out your result percentages and flipping that same coin four-thousand times to get your result percentages. Four flips of the coin is an inadequate sample size - you may end up with three of the four or even all four of your coin flips coming up with the same result - to allow proper tweaking of numbers to truly reflect where things are versus where they should be. Four-thousand flips of the coin will result in a far more accurate representation of the true contrast between results as it is highly unlikely that you will end up with that coin turning up the same result four-thousand times in a row.

I'm just using coin flipping as an incredibly basic example for the sake of ease of understanding. In reality, there are far, far more variables that need to be considered in balancing classes and skills.

[Warlyx.6732]

it isnt like beta existed or alpha , because some elites are horrible and some OP as hell....hell some weapons in some professions are useless.

then we have the long wait btw patches , u nerf a spec a lot (PVP or PVE) and there u go , 3 months until your spec is maybe....if u are lucky retweaked a bit and got some love back ....other times u are still nerfed some more (cough power reaper cough)

[Daffan.8924]

if they designed around the idea of split values for pvp and pve, then it wouldn't take them 2000 years to work out a compromise between pvp and pve, because you wouldn't need one.

[Llethander.3972]

@Daffan.8924 said:if they designed around the idea of split values for pvp and pve, then it wouldn't take them 2000 years to work out a compromise between pvp and pve, because you wouldn't need one.

That's what they do. Many skills have different values based on whether the player is in PvE or PvP. They still have to wait for enough time to pass to actually have an adequate sample size of information in order to determine what changes need to be made.

[maddoctor.2738]

@Llethander.3972 said:

@Daffan.8924 said:if they designed around the idea of split values for pvp and pve, then it wouldn't take them 2000 years to work out a compromise between pvp and pve, because you wouldn't need one.

That's what they do. Many skills have different values based on whether the player is in PvE or PvP. They still have to wait for enough time to pass to actually have an adequate sample size of information in order to determine what changes need to be made.

The problem is they allowed untested and broken pof elite specs to be used in an ongoing pvp season. It would've been much better if the pof elite specs weren't allowed in ranked arena, until there was enough time to test them and balance them properly. Dropping a major bomb like that in the middle of a pvp season was a huge mistake.

[Llethander.3972]

@maddoctor.2738 said:

@Llethander.3972 said:

@Daffan.8924 said:if they designed around the idea of split values for pvp and pve, then it wouldn't take them 2000 years to work out a compromise between pvp and pve, because you wouldn't need one.

That's what they do. Many skills have different values based on whether the player is in PvE or PvP. They still have to wait for enough time to pass to actually have an adequate sample size of information in order to determine what changes need to be made.

The problem is they allowed untested and broken pof elite specs to be used in an ongoing pvp season. It would've been much better if the pof elite specs weren't allowed in ranked arena, until there was enough time to test them and balance them properly. Dropping a major bomb like that in the middle of a pvp season was a huge mistake.

Yeah, that's usually the case. Introducing new elements mid-season is typically frowned upon and for good reason.

[Bladestrom.6425]

@OriOri.8724 said:@vesica tempestas.1563 Continuous integration systems and balance patches are about as far away from each other as you can get while still calling it programming. You don't balance classes by making changes and making sure none of your test cases break. You balance them by toning down what is overpowered and buffing what is underpowered. A continuous integration system has nothing to do with that. Also, in your comment you conveniently ignore all of the other updates that come to the game besides balance patches. All in all pretty much your entire comment is just inaccurate

the ponit was you can deploy every day if your develomment pipeline is mature enough. If you can deploy at that pace you can make very tiny incremental changes, i.e if a class is behind the dps curve, give them o.oo5% increase today, then tomorrow, then the day after etc etc. The alternative is relatively big class/skill stat swings that become risky, which means a lot of analysis, therefore taking longer, therefore not keeping up with the meta play and emergent issues.

[Einlanzer.1627]

The small team is also a problem because it essentially means they don't get broad enough input for needed changes, and as a result there are a lot of very ham-fisted and/or myopic balance updates and no updates for things that need them for ridiculously long stretches of time.

Recently, there was the change to the Guardian staff, which made no sense considering the mid-range effect on #1 was its only real selling point. Lots of weapons have dramatically undertuned #1 skills which has the effect of making them highly situational. See: dual Shot (Warrior Longbow - fires way too slowly for its damage) ,Elementalist staff (other than Fireball) - Water, Earth, and Air are all laughably under-tuned, which pushes ele players into spending most of their time camping Fire. It's absolutely silly that we go patch after patch without changes to these things.

[ugrakarma.9416]

I am adept at the idea that the more you try to concert, the worse it gets, so having only 1 balance every year is good. I hate having to keep reading skill change notes, and all this blah blah about new metas.

[Ithilwen.1529]

@OriOri.8724 said:Its because the balance team is too small. They have 3(?) devs on it, and these devs are also in charge of designing new elite specs for the expansions, which are coming out ~every 2 years. Not to mention they have to balance for 2 different game modes ( and should be balancing for 3 game modes, WvW needs its own balancing). 3 devs balancing 9 classes, 18 elite specializations, for 2.5 game modes, on top of having to design 9 new elite specs every 1.5-2 years is just too much. They can't get balance patches out in time.

If you don't believe me, their most common answer in the last balance patch AMA (back on the old forums) was "We wanted to do X but we didn't have time ". They simply need more devs on this team. Double the number at least, and that would give us enough devs on the balance team to do full overhauls like they did for guard spirit weapons, mesmer phantasm attacks, mesmer scepter a while ago.

I thinks that's a big part of the issue. The same people designing the specs are attempting to balance them. There's a huge conflict of interest there. No wonder someone's pet spec stays OP for long periods. There needs to be a separation between design and balance. Many, if not most, people will struggle to be objective about there own work.